DAW Fans: Try This

[Anderton]

Okay, let's rally the SSS product test brigade, and let me know what you think.

 

I have a dual ADAT optical interface. For quite a while now, I've assigned track outputs directly to ADAT outs, which interface with a Panasonic DA7. I then do my mixing within the DA7 (although most automation and submixing is handled in the DAW program).

 

If I assign those same tracks to a master stereo out within the DAW, then monitor that single stereo out through the DAT, it sure seems that the sound just isn't quite as good. Am I going crazy, or have other people noticed this kind of sonic degradation too? Now, part of this could be that I often use the DA7's EQ and dynamics, which takes a load off the DAW's CPU. But I suspect there might be more to it than this. Any thoughts?!?

[mattzen]

A lot of pros prefer not to do the summing in the DAW. Direct outs to a digital board always seem to be better. I try to mix as little as possible in my DAW. Direct outs to my o2R do sound better to me. I would also do a test summing all of your outs to a stero pair on the DA7. This will take the DATs converters out of the equation. They can be contributing to the difference you hear. Just a thought.

 

There has been some discussion of this on Roger's board and George's board. I don't remember the particular threads but a search should yeild something. If anyone remembers the thread....

[Anderton]

<>

 

They're already out of the equation. When mixing to the DA7, I just go in through the ADAT interface. When mixing within the DAW and sending an output to the DA7, I just use one stereo pair of the ADAT interface. The only difference is that the mixing is done in the DA7 instead of the computer, and it doesn't hit analog until the inputs of my 20/20 active monitors.

[Philip OKeefe]

Craig, I have a Dakota / Montana setup too (thanks for the tip on those cards!), except mine are running the ADAT lightpipes to a pair of AW4416's. I've noticed the exact same thing when I sum to a stereo pair - it sounds better going to individual channels. Logic Platinum 4.7 seems to sound a bit worse in this respect than SF Vegas Audio 2.0. I can't say for certain why this would be the case, but maybe it's because Logic's a more complex program with more overhead...? For the record: My system is a PIII 933 MHz, 384 MB RAM and dual W.D. 7200 RPM hard disks. Not exactly "state of the art" but not a low end slouch either...

 

In my tests, I ran the summed tracks sans any plug ins or automation whatsoever on the DAW software, running through a stereo pair on one of the lightpipe cables into the AW4416, and then with the same parameters with the only difference being that the individual tracks were run to individual faders on the AW's instead of being summed in software to a stereo pair.

 

 

Phil O'Keefe

Sound Sanctuary Recording

Riverside CA

http://members.aol.com/ssanctuary/index.html

pokeefe777@msn.com

[lawrence_dup1]

Yep, yeah, uh-huh...

 

I discovered years ago that there is a definitive degredation in the sound of my daw (VST) when compared to my console (d8b) or even a good stereo editor (Wavelab). Summing in the daw seems to yeild a "smaller" and less accurate sound, less pleasing to my ear. This has not improved with subsequent daw upgrades.

 

Even before I bought the d8b I had resigned myself to the fact that mixing within the daw was not gonna cut it for me. No Craig...you are not bonkers.

 

I can't speak for Pro Tools or many other high end daws but I would never consider mixing a paying clients material with VST. With that said, the quality of the individual tracks (minus daw processing) rival anything out there. What goes in through my d8b -> 2408 lightpipe -> to VST comes out wonderfully full and rich, so much so that I've never considered a stand-alone HD recorder until I heard of the Alesis vaporware product.

 

Lawrence

[Curve Dominant]

Yo, waahssup, guys...

 

For the benefit of folks like myself who have yet to use these tools, but plan to, and would like to follow this thread, could someone pleez breakdown some of these terms and gear configurations in more detail.

 

Specifically:

 

the difference between "summing" in the DAW, and mixing DAW tracks on an external board

are you talking about sending each individual track from the DAW to the external board, or just left and right channels

what options concerning hardwiring such a configuration are the most efficient and cost-effective

 

I'm considering Digi001/PTLE as an expansion of my Roland VS, so this subject caught my attention, but yet my ignorance of the technical details has kept this discussion well over my head. I applaud you in advance for enlightening me, and others who may benefit as well.

 

CLAP! CLAP! CLAP! CLAP! CLAP! CLAP! CLAP! CLAP! CLAP! CLAP! CLAP! CLAP! CLAP! CLAP! CLAP! CLAP! CLAP! CLAP! CLAP! CLAP! CLAP! CLAP! CLAP! CLAP! CLAP! CLAP! CLAP! CLAP! CLAP! CLAP! CLAP! CLAP! CLAP! CLAP! CLAP! CLAP! CLAP! CLAP! CLAP! CLAP! CLAP! CLAP! CLAP! CLAP! CLAP! CLAP! CLAP! CLAP!

 

le curve

[alphajerk]

thats funny because stevepow did a test of several DAWs and an O2R digitally summing the channels both ways [sans fx/automation] and the output files are EXACTLY the same... hmmmm.

[Hippie]

I have noticed this same, "thinning" of sound", from the DAW versus straight off the ADAT'S. I don't know what exactly the problem is, but I'll take a guess.

 

When I transfer recorded tracks from my 2 ADAT's to my DAW/Cakewalk via 2 lightpipe interfaces, (MOTU 2408), the record meters in Cakewalk, at the "default nominal level" of 0db", are reading much hotter than the ADAT's level meters read. (If all things "digital" are equal, while using a "0db" record level in the DAW, there should be NO gain changes from the original, this is NOT the case!).... So, I have to compensate by "fadering down" the DAWs' incoming signal to a level similar to that of the ADAT's level, which in turn alters the gain structure of what was originally recorded, causing them to sound different in the final result/mixdown.

 

Transferring digital signals from the ADAT's to the DAW can be pretty dicey, and often takes me several passes to get the gain structures equal or close to the original; .. It's a tricky maneuver to capture the full original signal into the DAW without altering the signal thru gain changes to be any hotter or colder, otherwise I wind up with something that is not quite the same as the original- (too cold goin' in to the DAW, will make it sound thinner, too hot & I'm screwed, 'cuz there will be distortion and/or no headroom left for EQing or plug-ins that will hotten the signal further.) -It can be sort of a "guessing game" of where the levels need to be during the transfer into the DAW, and what alterations to the signal will be forthcoming in the DAW, to do the least harm in comparison to the original.

 

Apparently, the MOTU-2408 has a more sensitive optical input than the ADAT's optical output, otherwise, in Cakewalk, I would be able to do a digital transfer from the ADAT's with all the faders in Cakewalk at a 0 db level, for an exact dupe, right?

This tells me that there is some variation of the "ADAT optical interface protocol" between different devices using them, at least in regards to levels.

 

-Hippie

[Alndln]

I noticed the difference when I used to mix "out", but Iv'e been getting such good results staying "in" that I don't have to think about it anymore.Although in Nuendo the same problem seems to exist internally and externally,so I treat my tracks there(EQ, effects ect.)and export the tracks individually back to Sonar/Cakewalk,then do an internal mix straight to stereo wave and get exactly what I put in.In my situation I started thinking about how many stages of conversion and EQ my stuff goes through before the final product and decided the old way (mixing out) was way too many.Think about it,Mixing board=and or pre=Adat/Roland vs=Daw,Daw=mixing board/pre=adat/Roland/dat. No thanks,I'm staying in.I can get the same exact thing going in as going out(warm pre sound ect.),but it requires getting it right going in,something I learned from the old day's when the clock was running.

[Philip OKeefe]

Originally posted by Curve Dominant:

For the benefit of folks like myself who have yet to use these tools, but plan to, and would like to follow this thread, could someone pleez breakdown some of these terms and gear configurations in more detail.

 

Specifically:

 

the difference between "summing" in the DAW, and mixing DAW tracks on an external board

are you talking about sending each individual track from the DAW to the external board, or just left and right channels

 

 

Say you have eight tracks recorded on your computer DAW. All eight parts are on their own DAW tracks. Now the DAW has to have a way to get the sounds OUT of the box and routed somewhere else, right? If you only have a stereo out soundcard (say, a Soundmasher, etc.) when you play back those tracks, they're being mixed from 8 down to two.

 

On higher end soundcards, you will usually find multiple outputs. For example, PT LE with a Digi 001 has several analog I/O's as well as S/PDIF I/O and an 8 channel ADAT lightpipe I/O. Say we're using the 8 channel lightpipe output from the computer DAW to route the tracks over to our digital mixer. In the software, you can assign which physical output "channel" the software will route the audio to from any particular track. I could send the info on DAW track 1 out via lightpipe channel 8 if I wanted to. Each DAW track can be sent out on a seperate lightpipe channel or several of them can be summed and SHARE an output channel.

 

What we've been discussing is that some of us feel things sound better when we do NOT sum channels in the computer, but rather send them out individual outputs into our digital mixers and do the summing over there.

 

 

what options concerning hardwiring such a configuration are the most efficient and cost-effective

 

I'm considering Digi001/PTLE as an expansion of my Roland VS, so this subject caught my attention, but yet my ignorance of the technical details has kept this discussion well over my head. I applaud you in advance for enlightening me, and others who may benefit as well.

 

 

Curve my man, hardwiring anything's never the best idea out there... with all the lightpipe channels I have around here, I'm currently looking for a lightpipe patchbay... Although if you're just using PT LE and a Digi 001, it shouldn't be too hard to connect everything.

 

The main drawback as I see it is that your Roland doesn't have any way that I know of to get ADAT lightpipe INTO it... unless there's some sort of lightpipe to S/PDIF converter... and what's the point of that? Just use the Digi 001's S/PDIF output. No matter how you slice it, you're going to be stuck transferring summed parts via S/PDIF if your mixing board doesn't have the capabuility to accept multi-channel digital inputs.

 

Your only other option is to use the analog outputs on the Digi 001, route them to some sort of analog submixer and / or individual analog inputs on your VS890. Then you're going to get extra signal degradation from the extra D/A and A/D conversions, as well as any noise from the submixer, etc. Not optimal either. I know how much you dig your Roland - Maybe it's time to consider a VS1680, 1880 or even one of those hot new VS2480's? Or, you might consider getting the new Tascam DM-24 digital mixer (shipping this summer) or maybe one of the other digital boards that take option cards with ADAT lightpipe I/O on it if you're really set on going with PT. Yamaha just announced a "Jr" version of the AW4416 - it only has one option card slot, but would make a nice match with Pro Tools LE for some users. You can check it out at :

 

http://www.aw2816.com

 

 

Consider your needs and what you really want / need to accomplish, then design the system *first* before plunking down any hard earned cash. you'll thank yourself later, trust me...

 

Phil O'Keefe

Sound Sanctuary Recording

Riverside CA

http://members.aol.com/ssanctuary/index.html

email: pokeefe777@msn.com

[Mr Darling]

I have to say Craig that I've never notice this and it scares me to read it. How can you explian this? and what about people who mix all in protools without any external mixers? No, I can't deal with this info so I'll ignor it.... http://www.musicplayer.com/ubb/confused.gif better continue to work the same way and pretend everything is good http://www.musicplayer.com/ubb/biggrin.gif

 

for Curve what the are talking about is : let's say you have 4 vocals tracks. If you send each vocals seperate to the mixer that mixing on external , if you send those 4 vocals via a stero track (pan and level set in the DAW) that's summing.

 

 

 

------------------

Visit http://www.DarlingNikkie.com/sounds for free MP3's

[Curve Dominant]

Yo, thanks Phil! You're a good man, taking the time to break it down for me like that. It makes sense to me now. AAHHH...KNOWLEDGE! Now, concerning the following:

 

>>Say we're using the 8 channel lightpipe output from the computer DAW to route the tracks over to our digital mixer. In the software, you can assign which physical output "channel" the software will route the audio to from any particular track. I could send the info on DAW track 1 out via lightpipe channel 8 if I wanted to. Each DAW track can be sent out on a seperate lightpipe channel or several of them can be summed and SHARE an output channel.

What we've been discussing is that some of us feel things sound better when we do NOT sum channels in the computer, but rather send them out individual outputs into our digital mixers and do the summing over there.<<

 

What I'm now curious about is: does the sound degrade progressively the more you sum? In otherwords, say I had a 16-track recording on my (still imaginary) Digi001/PTLE rig. Since the lightpipe is only 8 channels, I'll combine the tracks into pairs, to create 8 channels, for mixing on an 8-channel digital board. Then, I would mix the same 16 tracks again, but this time I would sum all 16 tracks to one stereo pair. Would I notice the the degradation in that first scenario less than I would if I had summed all 16 tracks into one stereo pair? (Does that make sense?)

 

>>The main drawback as I see it is that your Roland doesn't have any way that I know of to get ADAT lightpipe INTO it...<<

 

Yeah, that nixes my idea of using the VS as my outboard mixer. It does have an optical input, but it's a track input, not a 8-channel ADAT input. Oh well! No worries, I don't even have Digi001 yet, but it's good to know ahead of time that researching an outboard mixing option will be in order, so thanks for the "heads-up" on that, guys!

 

Mr. Darling, don't be too discouraged. I thought getting Digi001 was going to complete my rig, so I wasn't happy with the news either, but this is a tough racket. But look at it this way: I've done a couple of projects in a Digital Performer studio where we mixed down to stereo from DP without the benefit of an outboard mixer, and they were two of my biggest resumé builders. Yeah, they probably would have sounded phatter if we didn't sum the tracks, but who knew? We're blessed to have guys like Craig constantly pushing the quality envelope forward for us, but if you're not there yet, don't let it bust your groove, man. "Sugar Cane" is still phatt.

 

Ever grateful,

curvedominant

[Anderton]

<>

 

What I've noticed is a sort of intermodulation distortion in the low end. Now that people have confirmed that I'm not crazy, and they've noticed a similar phenomenon, I'm trying to come up with theories. The fact that some people HAVEN'T noticed this problem is also encouraging. In THEORY, there shouldn't be any difference, but apparently in practice, something can screw up the mixing process.

 

I'm thinking that the problem may be due to cumulative DC offsets throwing off the math in the DAW's summing process. The fact that this seems to be low frequency oriented tends to support that -- I may be getting some subsonics because I haven't done my usual "get rid of anything under 15 Hz processing" yet in the tune where I noticed this.

 

Another possibility is that I'm overloading my CPU. The tune is hitting about 70%, and maybe the DAW is missing a few critical cycles here and there. I'm going to try mixing down the soft synths etc. to audio tracks, lighten the load on the CPU, and see if I still notice the same issues.

 

Also, I'm doing this in 16 bit. I'm curious to see if 24 bit files exhibit the same issues.

 

Thanks for your replies,. I'm definitely interested in pursuing this further...by the way I haven't seen this restricted to one particular program, and the fact that some of you mentioned Logic and VST exhibiting the same problem is interesting. It seems it is the summing process itself, not the program, where the problem occurs. But again, this is all speculation...............

[Dylan]

Craig,

 

Have you put digital clock into the equation? It could be that the clock from the software/Dakota that feeds into the DAT isn't stable. Do you have an audio card with an analog output that you can test with? I mix everything inside my DAW with Vegas, and I haven't noticed any loss in audio quality when I mixdown. But then again, I generally don't add a ton of plug-ins to my mixes, so this might have something to do with it. I generally work at 24-bit as well. I have noticed that my mixes done in Vegas tend to sound more open than they do in Pro Audio 9. Sonar's mixing engine supposedly sounds significantly better, but I'm holding off using Sonar until they work out some more bugs. I guess there have been complaints about the summing bus in ProTools as well recently. My best mixes have been done with external mixers (both digital and analog), even when I have had to convert from digital to analog out of the audio card and back to digital into the digital mixer when a direct digital connection wasn't possible.

 

-Dylan

 

This message has been edited by Dylan Walters on 06-19-2001 at 02:19 PM

[D.Triny]

another item worth investigating is the pan law

 

-david abraham

[Logan]

The biggest sonic hit/degradation that I notice is the one that I get when I go to 16 bit. I have been trying to mix inside the box,(while trying to think outside the box ;-) sorry couldn't resist, feel free to slap me if I degenerate into physco babble again) that is, inside my Daw ( dual p111 800 fast Scsi drives running Nuendo), because it just seems to make sense not to go through too many conversions. The summed mix at 24/48 of say 20 tracks is okay and has some punch and separation, but man when it's dithered down (Apogee UV22) it's a mere shadow of itself, and to date nothing touches the 16 track analogue machine on rhythm sections (guess I'm a sucker for tape compression)

I've been following this topic on a number of boards and there are significant numbers of folks who are complaining about the summing buss on virtually every recording program that exists. On the Nuendo board some folks are saying Logic sounds better and Grand pooba Charlie Steinberg chimed in and attributed the difference to the value of the pan law which apparently is 0db in Logic, so some mixes sound louder. We all know that louder is better ;-). I don't know 'cause I've never used Logic.

I still think Wavelab sounds better than most other programs but Steve Pow's test seem to deny that.

Some folks say that the more tracks you sum the worse the degradation is, which would give credibility to Craig's DC offset theory or CPU overload, causing lackadaisical math (can this happen?)

I have a couple of 12 -16 track projects on tape (not earth shattering stuff but good for a test), so I'm going to load one into the DAW and do a mix inside the DAW and one sending 16 channels back to an analogue board (Soundcraft Series 800) then recorded to Wavelab. I'll also do a mix inside the DAW and send each track through a DC offset correction plug. I'll do an analogue mix from the tape deck and record the stereo buss to Wavelab. So I'll have four mixes to evaluate, Won't be the same as using a digi board but it should be interesting. I wish I could do a mix from the DAW separating each channel through the board to a two track analogue machine , because I suspect this is the way to go, unfortunately the two track is no longer in the building. I'll let you know what I think and maybe I can burn a few copies and send it to some of you folks to see what you think. I'll try and get at it on the weekend as I still have to finish the project I'm working on, which doesn't have enough tracks to test the theory of more tracks 'causing the problem. take care Logan

[Anderton]

<>

 

I think I missed something. What test?

[alphajerk]

craig, it was a test the stevepow did testing DAW vs. Analog boards summing

 

email i recieved:

"Check it out - all the digital mixes I did are the identical wav form - all

the stereo mix wav files are identical!!!! I couldn't hear anything

different at all after wasting hours listening to them, so I pulled them

into Samplitude and compared them two at a time - lined them up sample for

sample and then flipped the phase on one pair. Dead silence - perfect

cancellation. Even the O2R and Mixtreme Hardware Card mixes were the same

as Cakewalk, Cubase, Vegas, and Samplitude. Any pair mixed together would

cancel out completely with the phase flipped on one of them. Of course, the

Allen & Heath analog mix was different - but very, very close.

 

Whaddaya make o' that?"

 

i made some stunned like a deer in headlights comment.

 

 

"Yep - I almost fell over as one by one, they were all identical - even the

O2R and Mixtreme. I suppose EQ, Compression and Plug-ins are what

differentiate the final product. I'm betting PT would come out the same as

well. So - you thinking what Roger said might be true after all?

 

It was also pretty wild how close the A&H came to matching up - I could

easily see the same waveform samples to line them up. That was with me just

setting the faders and input pots to ZERO by eye - no calibration tests or

anything. "

 

it was due to an arguement that i had with eric sarafin over analog vs. digital summing. basically he claimed that even the cheapest analog boards sound FAR better than any digital. i dispute that claiming that digital mixing is better than all but the high end analog boards [considering that your source is digital going out individual outputs to a mixer vs. the stereo feed of the internat digital mixer] and only then are they beneficial to more easily implement great analog outboard [not necessarily expensive, just great]

[Curve Dominant]

I vaguely remember Steve posting that. I also remember seeing posts where some claim that certain DAW software systems sounded better than others. I dunno, this is not a scientific opinion, but I was always a little skeptical of such claims. It just doesn't seem to make sense at a gut-level. The various softwares (DP,PT, Cubase, Logic, etc...) are different in a UI sense, of course, features, working environment. But sound quality? It's all ones and zeroes, no? Especially if the bit-depth and sampling rate is constant.

 

>> part of this could be that I often use the DA7's EQ and dynamics, which takes a load off the DAW's CPU.<<

 

That would seem to be a variable. Maybe we should get Paul Frindle over here. Paul?...

 

curvedominant

[seclusion]

I run Logic, 4.7, 2 tascam tmd4000 consoles, a dedicated gigastudio puter(2408), and the other running logic(2408)...

Wow this is very deep...

This hole process of summing really has been driving me nuts even before when I was mixing from a mackie 8bus analog out to a cheep fostex d5..

I'd mix and mix and everytime I listened to the dat I'd wanna puke.. So I tried going back into the Daw unbalanced(orig 2408) and mix it to logic..

To me it improved from the Dat but I still didn't like it..

So now I'm Aes out,digi board to the Dat(No D5 converters).. A huge improvement again listening to the mix then the dat and back... Now I dither through the 4000, the 24 bit stuff down to 16(SAW) and still notice a "Change"..

What I do notice is that if I send that dat Aes back in through the board st to logic again( so I can Burn it) it sounds different again.. I open up a new song with nothin else(eq/plugs) going.. I doin wordclock from a MOTU midi timepeice Av seems very stable.

Now I haven't ever summed all the daw stuff together internal with out a board cause I'd go nuts..

But I then tried to mix 20 audio tracks, midi, etc. stereo out of the board to logic(24bit) This seems the best "sound" to me.. If I ab the logic st track and the dat, logic has it beat.. But then I'm into 24 versus the 16 bit thing and I don't wanna go there..

So you're sayin getting that mackie/logic controller thingy and bypassing the digi boards and totally mixing in a Daw may NOT be a good thing..

Thanx again for the heads up..

Brian

[Gordon_dup2]

" . . . inside my Daw ( dual p111 800 fast Scsi drives running Nuendo), because it just seems to make sense not to go through too many conversions. " - Logan

 

Way off topic. What motherboard are you using for your dual P-III system?

 

BTW - I notice a deffinate improvement in sound quality when I use my Yamaha DSP factory channels over the summed-VST mixer. The tracks seem to have much more depth and seperation.

[Logan]

Gordon

I'm using a MSI 694D pro and I'm running very stable, however I don't use midi in this machine and this is significant, because the Via Chip sets have problems(some say) with midi. Check out the RME site for discussions and recommendations on this, they are very good at testing and recommending systems. There is a fix for this board that involves soldering a resistor, to correct midi timing but I have enough trouble soldering cables with out messing with my mobo, so I've never tried it. I use a different machine when I need midi then record the tracks as audio into the audio project I'm working on.

Back on topic: Someone ,maybe Curve, mentioned that the comparisons mentioned here are not scientific, I agree but I don't see how this will ever be achieved there are just to many different factors involved. which is why I'm doing the comparison of the tracks summed in the software and the tracks mixed through an analogue board. Reality is that this is probably the configuration that most folks are using, a DAW whether computer based or stand alone, coupled with and analogue board. I went right from tape based to DAW and started working right away, mixing in the box and never really did the tests I should have to decide which way of mixing provides the best product for the folks I record. Now folks seem happy with the work, but all my DAW projects have been solo classical guitar or a piano duo or a solo blues guitarist/singer, so I haven't had a number of tracks to jam up the mixing buss, but I've never been satisfied yet.So the issue with me is does the mix sound better avoiding the summing buss and going through the additional conversions. Hope to rectify that by doing a few days of tests. Take care Logan

[Anderton]

Of course, one issue is that even the slightest level difference screws up the comparison. People listening to two identical sources, but one as little as 0.2 dB louder than the other, will say the louder one has better imaging, or warmth, or whatever.

 

I'm going to run some more tests. Last night, for some reason I can only dimly understand, the clock source on my card changed to INT from DIG IN. (I didn't change it on purpose, so who knows?) Interestingly, this didn't make an obvious difference on playback, but did add gaps while recording -- which is why I thought to check the clock.

 

It's possible that I had different clock settings when mixing internally compared to mixing through the DA7 mixer. So it's back to the test bench.

[seclusion]

Don't ya love it...

I used to be so confident mixing to adat and dat and feel good knowing it sounded so much clearer than the portastudio...

But damn does this quest for "Better" ever go away..

Guess not..

Brian

[Lee Flier]

The problem I had with Steve's test is that it had to be done without any processing in order to keep the criteria consistent between different DAW systems. I think whatever happens in summing that sucks (and yes, I've noticed it too, big time), the amount of DSP that you've done is a big part of it. Level changes, EQ, plugins, etc., I think they all contribute to whatever gets screwed up when you sum. Of course there is no way to totally objectively compare different systems at that point because the DSP is all different between them. But I think enough people have noticed this degradation that we should be past wondering "if" there is really a problem and move on to "what" the problem is.

 

And yes it has been discussed on other forums, but finding the time and energy to do exhaustive testing is tough.

 

--Lee

[Logan]

Lee

I suspect you are right in your ascertion that it's time to get into the discussion of what is wrong and not is something wrong. However, because, as I said, I moved over to working in the Daw from analogue tape and I have attributed all the problems I've had (not liking the digi sound as much as the tape sound), to the difference in digi and analogue formats, I feel I must do some testing to see if the summing buss is the problem. I also think that your idea that it's the DSP level involved in the tracks is perceptive and there are others beating the same drum in this discussion, around different boards. I'm not all that concerned with a totally scientific test just the best mix. So if it is a DSP issue/ channel then I will also try running the tracks from the DAW without the plugs and EQ and do that in the board. My limit will be compression as I've grown perhaps overly ;-)fond of the Ren Comp and I only have 4 hardware comps to throw into the mix and none of them are optos. The Eq in my board is great so I expect improvement there and I think that my hardware Revs will stand up to the Waves and TC software verbs I'm using now. The whole deal is that unfortunately I think I settled for a sound(ALWAYS A MISTAKE) and this thread has put the fire in my belly to re-explore all the options.

I moved to DAW for two reasons 1: a gear split with a studio partner left me without access to a 1" tape machine and I was faced with using the 1/2 " B 16, so I looked long and hard at a 2" machine, but reality was 2: the bands I work with could not afford the costs of working with 2" tape and barely could afford 1" tape, hell 30 mins of 1/2" tape here in Canada is $70. So I looked at DAW as a way to cut initial costs and editing costs and to increase my income on the tracking and mixing side. The bands I deal with are having a hard time existing, as costs for gear and transportation are much greater here in Canada then states side, and venues are few and far between, so bottom line is that I really want this DAW stuff to work. Take care Logan

[Curve Dominant]

First off, great thread, Craig. This one will no doubt go down in the SSS hall of fame when the smoke clears.

 

Logan posted, among other things:

>>Someone ,maybe Curve, mentioned that the comparisons mentioned here are not scientific, I agree but I don't see how this will ever be achieved there are just to many different factors involved.<<

 

Logan: I am a composer, not a scientist, so please do not interpret my input as a qualified scientific opinion. What I had posted was:

>> part of this could be that I often use the DA7's EQ and dynamics, which takes a load off the DAW's CPU.<<

 

That would seem to be a variable.

 

That was simply an observation of the possible consequences of Craig (who's words are italicized above) using the DSP in the outboard mixer, rather than the DAW; as such it would take a load off of the CPU in Craig's DAW, which may indicate a more efficient use of processing power, and therefore a better-sounding result over-all.

 

Lee picked up on this, with:

>>I think whatever happens in summing that sucks (and yes, I've noticed it too, big time), the amount of DSP that you've done is a big part of it. Level changes, EQ, plugins, etc., I think they all contribute to whatever gets screwed up when you sum.<<

 

Now, even for someone who does not own a DAW (not yet, anyway), that statement seems completely logical. Just knowing how computers as machinery operate, it makes sense. I don't think there ever was a question of "if" there is a problem. What I am interested in is:

1. When is it a problem, or at what point, or under what circumstances does it become a problem?

2. When (or under what circumstances and conditions) is it not a problem, and what can that tell us?

3. Does the frequency and/or amplitude of the problem shift as circumstances and conditions shift? If so, are the shifts corallary?

4. What changes in circumstances and conditions seem to alleviate the problem as it exists?

 

Now, this is interesting, so take note: I was at Turtle Studios last Friday night for a Philadelphia Music Conference showcase. At one point while I was chillin' in the control room with a glass of Merlot, Jay Levin (Turtle partner and resident computer engineer) sat down next to me and said, "Check this out, I put a PowerCore in the G4." Instantly seeming to sense my cluelessness (Jay is a very observant chap), he explained that the PowerCore is this chumpy that is installed in a PCI slot, and it takes ALL of the DSP processing load OFF of the native CPU on a computer that's running a DAW. He told me that it was just announced at Namm this past spring, and he'd been dying to snag one.

 

So my question is: are any of you folks using such a component? Have you heard about this thang?

 

Well, anyway, I tell ya what I'm-a gonna do: I'll email Jay with a link to this thread. Maybe, if he's not too busy, I can coax him into posting the results of this jimmy-jam he's using, which would seem to offer some relief to all the poor lil' ol' CPU's out there that are feelin' the heat and dyin' for an ice-cold glass of digital lemonade.

 

In the meantime, here's what I've learned from all this: when I do get the Digi001/PTLE rig, I'm gonna hold onto the Roland VS880EX and track to that, EQ/FX/ETC in it, and then export those tracks to PTLE, which I'll use just for editing and file-formatting. Sound good?

 

curvedominant

 

 

 

This message has been edited by Curve Dominant on 06-21-2001 at 01:24 AM

[mattzen]

Hey Curve, it called the TC PowerCore and it's been discussed on a couple of threads here's one:

http://www.musicplayer.com/ubb/Forum17/HTML/001099.html

It seems dissappointing and is definitely not the solution to our DAW summing problems.

 

BTW, watch out with the tracking to the Roland....I believe it uses data compression on your audio in every mode but mastering mode (I'm sure someone will correct me if I'm wrong). It has a sound to me and I don't like it. Also, I would hope that the converters and the rest of the analog section on the 001 are better than that on the 880. This may be another weak link. And surely I think that the EQ and FX in PT are better than those in the Roland. You are looking to upgrade here and using the roland that extensively with PTLE doesn't seem like much of an upgrade.

If you really want to take a load off of the processor while gaining benefit of plugs, bounce each track individually to disk with the plugs on them. Then remove the plugs active in real-time. I don't know if it will clear up the summing issue at all but it will take the load off of your processor.

 

Don't get me wrong, roland makes nice all-in-one boxes and you still will have use for it when you get PTLE. For example, I'm involved in a project that uses one for location recording.

 

When you get your new gear you will develop a new style of working ...you will figure out what works the best and what sounds the best to YOU.

[Curve Dominant]

Posted by mattzen, among other things:

Hey Curve, it called the TC PowerCore and it's been discussed on a couple of threads here's one:

http://www.musicplayer.com/ubb/Forum17/HTML/001099.html

It seems dissappointing and is definitely not the solution to our DAW summing problems.

 

And after a complete reading, here is the most definitive post that I could find on that thread:

Greetings to all of you!

I am the Andrew who was working with Nika over the weekend on the Powercore tests. I just wanted to give a little anecdote about the testing that I thought was appropriate (Nika has done a great job of presenting the facts).

The 40 track project we worked up started off as about 30 and grew. I wrote the project average to large to test CPU vs track count capacity. The piece was originally saved on a firewire hard drive and then transferred to an ultrawide SCSI drive connected to an ATTO card.

 

Originally the piece was written on a G4/500. My personal computer is a G3/266 with 320 Megs of RAM, ATTO card, and UW SCSI drive. I wasn't planning on keeping the song as it was just for testing, but it turned out to be interesting, so I decided to try to finish it on the G3. I was delighted to be able to open and run the project on the G3. CPU usage was high (80%), but still running. I was short a couple of effects though - I couldn't open any native reverbs and the vocal and drums really needed something (I already had 5 DP effects (2 EQ's, 2 Pre Amp1's, and a compressor) open). I installed the Powercore and was able to get 3 reverbs and an EQ. It was just what I needed.

 

To me, being able to finish a stupidly big song like that without having to buy a new computer, a better audio interface (so I could connect multiple outboard reverbs), or sacrifice the effects I wanted was great.

 

Most of the DP users up here are running faster machines than I am, so the issue may be a little different, but assuming that TC will improve the cards "shuttling" capability and that native manufacturers will begin to write for the card to me means that I can keep more of my money by not buying a new computer, outboard processor, and audio interface, and still have a decent sounding/functional system for composition and recording/mixing.

 

Truthfully, I was not entirely sold on the card until I used it on the older Mac. The cost vs benefit of EQ's was not as great as I had hoped on a fast machine. Having now seen the card and its plug ins in action in a mix where I was out of CPU power, I have completely changed my mind. Very useful, and that usefulness in the clutch could apply to any CPU speed and any Mac.

 

Thanks everybody.

-Andrew

 

I dunno, mattzen. Andrew hardly seemed disappointed. Besides, the issue remains whether a solution exists regarding a unit such as the Powercore, rather than the degree of your endorsement of a specific product. I won't even bother to get into it with you over how the VS works - I've used it for over a year now, and I'm fully aware of the compression options on that unit, the quality of it's DSP's and EQ's, etc, but thanks anyway for playing mom.

 

In otherwords, let's get back on topic: what's the solution to clean track summing within DAW's?

[Logan]

There are a few more dedicated DSP plugin cards available. One from folks who make the Pulsar card and one called a UA, I think I read that Yamaha has one in the works, but it may be an early morning delusion. There is some discussion of a couple of these over on Steinbergs Nuendo Forumn. They interest me 'cause some highend plugs like George's stuff may be able to be incorporated in a way that will deal with his fear that the plugs will be cracked. This will give access to us PC guys that we have not yet had. Also they may help deal with the DSP/track issue that we have been discussing. However like all new product releases it may be best to let the bugs sort themselves out. Take care Logan